Ruleplus-exception model of classification learning

The authors propose a rule-plus-exception model (RULEX) of classification learning. According to RULEX, people learn to classify objects by forming simple logical rules and remembering occa-sional exceptions to those rules. Because the learning process in RULEX is stochastic, the model predicts that individual Ss will vary greatly in the particular rules that are formed and the exceptions that are stored. Averaged classification data are presumed to represent mixtures of these highly idiosyncratic rules and exceptions. RULEX accounts for numerous fundamental classification phe-nomena, including prototype and specific exemplar effects, sensitivity to correlational information, difficulty of learning linearly separable versus nonlinearly separable categories, selective attention effects, and difficulty of learning concepts with rules of differing complexity. RULEX also predicts distributions of generalization patterns observed at the individual subject level. Psychologists have witnessed a major shift in the study of cat-egory learning during the past few decades. Early research was dominated by the concept-identification paradigm, in which subjects learned well-defined categories structured according to simple logical rules. Owing to the influence of researchers such as Rosch (1973) and Posner and Keele (1968), interest shifted to more ill-defined categories as might be found in the natural world. For ill-defined categories, no simple logical rules exist for classifying objects, and the boundaries demarcating alternative categories are fuzzy. With the shift in emphasis from well-defined to ill-defined categories, there has also been a major shift in the types of models used for explaining classification learning. Early re-search was dominated by hypothesis-testing and rule-formatio